Commerce,Coalitions,and Factor Mobility September 2002 freeing up more of the factor it uses relatively inten- issue that affects relative commodity prices-will di- sively (the scarce factor)than is demanded elsewhere vide an economy into very different types of coalitions in the economy at existing prices.When factor prices if there is substantial variation in levels of interindustry adjust,returns to the scarce factor fall even further than factor mobility (see Appendix A for a formal,general- the price of the imported good,while returns to the equilibrium treatment) abundant factor rise even further than the price of the exported good.The interindustry mobility of the fac- tors assures that trade affects owners of each factor in EVIDENCE OF TRENDS IN FACTOR MOBILITY IN THE U.S.ECONOMY the same way no matter where they are employed in the economy.This is the insight that encouraged Rogowski (1989)to anticipate broad-based conflict among owners Measuring Interindustry Factor Mobility of land,labor,and capital in trade politics.2 Given the obvious importance of interindustry factor Very different results are generated by alternative mobility in determining the income distribution effects types of models (often referred to as"Ricardo-Viner" of trade(and,hence,the politics of trade),it is vexing,as models)in which one or more factor of production is as- Grossman and Levinsohn(1989)have pointed out,that sumed to be immobile between industries (Jones 1971; very few attempts have actually been made to assess Mussa 1974.1982).3 In these models,the returns to levels of mobility empirically.The most direct evidence "specific"factors are tied closely to the fortunes of the has been provided in work on industry wage differen- industry in which they are employed.Factors specific to tials(e.g.,Krueger and Summers 1988),the response of export industries receive a real increase in returns due stock-market returns to import price shocks(Grossman to trade,while those employed in import-competing in- and Levinsohn 1989),and prices in secondary markets dustries lose in real terms.4 Factor specificity thus drives for capital equipment (Ramey and Shapiro 1998).6 All a wedge between members of the same class employed these studies suggest significant factor specificity and in different industries.The implication is that political sizable industry rents in U.S.manufacturing in recent coalitions form along industry lines,and this has guided years,but we do not have a historical standard of ref- much of the empirical analysis in the"endogenous pol- erence with which to compare these findings. icy"literature in economics that relates variation in To compare levels of factor mobility in the U.S.econ- import barriers across industries to the relative polit- omy in different periods,I have examined the variation ical strength of different industry-based groups (e.g., between rates of return for factors employed in differ- Anderson 1980:Lavergne 1983). ent industries.This is simply an application of the "law The Stolper-Samuelson and Ricardo-Viner models of one price."If factors are highly mobile (i.e.,mov- examine extreme,or polar,cases,in which productive able),return differentials should be arbitraged away factors are either perfectly mobile or specific.5 Fac- by (actual or potential)factor movement.Smaller dif- tor mobility is better regarded as a continuous vari- ferentials in wages and profits across industries are thus able,affected by a range of economic,technological, indicators of higher levels of mobility.The magnitude of and political conditions.Allowing that factors can have the differentials will reflect the costs of moving factors varying degrees of interindustry mobility,the simple between industries,which are influenced by a range of prediction is that broad class-based political coalitions economic and political variables,including the speci- are more likely where factor mobility is high,while ficity of human and physical capital to particular firms narrow industry-based coalitions are more likely where and industries,any factor market regulations that affect mobility is low.The trade issue-and,in fact,any policy firm entry and exit and hiring and firing,any policies that assist relocation and retraining,and the costs of transportation and communication.Different versions 2 Classes are defined here simply in terms of factor ownership:Each of this type of measure have been used previously in factor class comprises those individuals well endowed with a factor relative to the economy as a whole.This definition allows for the fact a wide range of studies of labor and capital mobility. that individuals often own a mix of factors(Mayer 1984). 3 The original model was introduced independently by Jones(1971) and Samuelson(1971):The former christened it the"specific-factors' 6 Magee (1980)examined the "revealed preferences"of industry model,while the latter named it the "Ricardo-Viner"model. groups to make inferences about mobility in his much-cited study Again,the logic is straightforward:A decrease in the domestic of testimony by labor unions and management groups before the production of an imported good releases any mobile factors for em- House Ways and Means Committee on the Trade Act of 1974. ployment elsewhere in the economy and thus renders factors specific In contrast,a great deal of empirical work has been done on the to the import-competing industry less productive,driving down their interregional mobility of labor and capital in the American economy real returns.Returns on the mobile factor rise relative to the price of aimed explicitly at uncovering historical trends,with much of the the imported good but fall relative to the price of exports,so that the attention focused on the geographic integration of the markets for income effects of trade for owners of this factor depend on patterns labor and capital during the nineteenth century (e.g.,Coelho and of consumption. Shepherd 1976;Davis 1965;Lebergott 1964:Odell 1989;Rosenbloom S The bifurcation is generally considered unproblematic in the eco 1990). nomics literature:Specific-factors effects are regarded as important 8 On industry wage variance in recent years,see Dickens and Katz in the short term but not the long term (Caves et al.1990,146-49; 1987,Gibbons and Katz 1992,Katz and Summers 1989,and Krueger Krugman and Obstfeld 1988,81;Mussa 1974).It is simply assumed and Summers 1987,1988.Almost all the work on the geographic that,over time,all factors are perfectly mobile.But this ignores integration of U.S.labor and financial markets has focused upon politics:Factor owners do not just choose between accepting their regional differences in wages and interest rates,and rate-of return returns in one industry and moving to another,they can also lobby differentials have also been used to gauge the level of international to influence policy (and hence returns). capital mobility (e.g.,Frankel 1991). 594Commerce, Coalitions, and Factor Mobility freeing up more of the factor it uses relatively inten￾sively (the scarce factor) than is demanded elsewhere in the economy at existing prices. When factor prices adiust. returns to the scarce factor fall even further than thi price of the imported good, while returns to the abundant factor rise even further than the price of the exported good. The interindustry mobility of the fac￾tors assures that trade affects owners of each factor in the same way no matter where they are employed in the economy. This is the insight that encouraged Rogowski (1989) to anticipate broad-based conflict among owners of land, labor, and capital in trade politics.* Very different results are generated by alternative types of models (often referred to as "Ricardo-Viner" models) in which one or more factor of production is as￾sumed to be immobile between industries (Jones 1971; Mussa 1974. 198213 In these models. the returns to "specific" factors ire tied closely to the fortunes of the industry in which they are employed. Factors specific to export industries receive a real increase in returns due to trade, while those employed in import-competing in￾dustries lose in real terms.4 Factor specificity thus drives a wedge between members of the same class employed in different industries. The implication is that political coalitions form along industry lines, and this has guided much of the empirical analysis in the "endogenous pol￾icy" literature in economics that relates variation in import barriers across industries to the relative polit￾ical strength of different industry-based groups (e.g., Anderson 1980; Lavergne 1983). The Stolper-Samuelson and Ricardo-Viner models examine extreme, or polar, cases, in which productive factors are either perfectly mobile or ~pecific.~ Fac￾tor mobility is better regarded as a continuous vari￾able, affected by a range of economic, technological, and political conditions. Allowing that factors can have varying degrees of interindustry mobility, the simple prediction is that broad class-based political coalitions are more likely where factor mobility is high, while narrow industry-based coalitions are more likely where mobility is low. The trade issue-and, in fact, any policy Classes are defined here simply in terms of factor ownership: Each factor class comprises those individuals well endowed with a factor relative to the economy as a whole. This definition allows for the fact that individuals often own a mix of factors (Mayer 1984). The original model was introduced independently by Jones (1971) and Samuelson (1 971): The former christened it the "specific-factors" model. while the latter named it the "Ricardo-Viner" model. "gain, the logic is straightforward: A decrease in the domestic production of an imported good releases any mobile factors for em￾ployment elsewhere in the economy and thus renders factors specific to the import-competing industry less productive, driving down their real returns. Returns on the mobile factor rise relative to the price of the imported good but fall relative to the price of exports, so that the income effects of trade for owners of this factor depend on patterns of consumption. The bifurcation is generally considered unproblematic in the eco￾nomics literature: Specific-factors effects are regarded as important in the short term but not the long term (Caves et al. 1990. 14649; Krugman and Obstfeld 1988, 81; Mussa 1974). It is simply assumed that, over time, all factors are perfectly mobile. But this ignores politics: Factor owners do not just choose between accepting their returns in one industry and moving to another, they can also lobby to influence policy (and hence returns). September 2002 issue that affects relative commodity prices-will di￾vide an economy into very different types of coalitions if there is substantial variation in levels of interindustry factor mobility (see Appendix A for a formal, general￾equilibrium treatment). EVIDENCE OF TRENDS IN FACTOR MOBILITY IN THE U.S. ECONOMY Measuring Interindustry Factor Mobility Given the obvious importance of interindustry factor mobility in determining the income distribution effects of trade (and, hence, the politics of trade), it is vexing, as Grossman and Levinsohn (1989) have pointed out, that very few attempts have actually been made to assess levels of mobility empirically. The most direct evidence has been provided in work on industry wage differen￾tials (e.g., Krueger and Summers 1988), the response of stock-market returns to import price shocks (Grossman and Levinsohn 1989), and prices in secondary markets for capital equipment (Ramey and Shapiro 1998)~ All these studies suggest significant factor specificity and sizable industry rents in U.S. manufacturing in recent years, but we do not have a historical standard of ref￾erence with which to compare these findings7 To compare levels of factor mobility in the U.S. econ￾omy in different periods, I have examined the variation between rates of return for factors employed in differ￾ent industries. This is simply an application of the "law of one price." If factors are highly mobile (i.e., mov￾able), return differentials should be arbitraged away by (actual or potential) factor movement. Smaller dif￾ferentials in wages and profits across industries are thus indicators of higher levels of mobility. The magnitude of the differentials will reflect the costs of moving factors between industries, which are influenced by a range of economic and political variables, including the speci￾ficity of human and physical capital to particular firms and industries, any factor market regulations that affect firm entry and exit and hiring and firing, any policies that assist relocation and retraining, and the costs of transportation and communication. Different versions of this type of measure have been used previously in a wide range of studies of labor and capital m~bility.~ Magee (1980) examined the "revealed preferences" of industry groups to make inferences about mobility in his much-cited study of testimony by labor unions and management groups before the House Ways and Means Committee on the Trade Act of 1974. 'In contrast. a great deal of empirical work has been done on the interregional mobility of labor and capital in the American economy aimed explicitly at uncovering historical trends, with much of the attention focused on the geographic integration of the markets for labor and capital during the nineteenth century (e.g., Coelho and Shepherd 1976; Davis 1965: Lebergott 1964; Odell1989; Rosenbloom 1990). 'On industry wage variance in recent years, see Dickens and Katz 1987, Gibbons and Katz 1992, Katz and Summers 1989, and Krueger and Summers 1987, 1988. Almost all the work on the geographic integration of U.S. labor and financial markets has focused upon regional differences in wages and interest rates, and rate-of return differentials have also been used to gauge the level of international capital mobility (e.g.. Frankel 1991)